Summary The pilot of a Bell 212 HP helicopter (registration C-FWDV, serial number30973) was carrying out heli-skiing operations in the Blue River area of British Columbia. After taking off from the top of a glacier, at about 8000feet above sea level, the pilot made a downwind approach to land at a pick-up area at the toe of another glacier. When the helicopter was at about 150feet above ground level, and at about 30knots air speed, the pilot increased the collective pitch to slow his rate of descent, but the engines (Pratt Whitney Canada PT6T-3DF) did not respond. The low rotor rpm warning sounded and the rotor rpm decreased. The pilot lowered the collective and confirmed that the rpm beep was full up and the engine throttles were fully open. The pilot flew the helicopter toward a snow-covered, frozen lake. The sink rate could not be arrested as the rotor rpm had not recovered, and the helicopter landed hard, yawed right about 90 and remained upright. The deep snow absorbed some of the impact forces, but the helicopter was substantially damaged. After the landing, the rotor rpm appeared to start accelerating and the pilot shut the engines down immediately. The pilot, the only person on board, was not injured. Ce rapport est galement disponible en franais. Other Factual Information The helicopter had incorporated Supplemental Type Certificates, which relate to the EnginePT6T-3DF installation. In this application, the Twin-Pac is not fitted with Pg air accumulators. The outside air temperature was between 0C and -5C and the emergency landing altitude was about 6100feet above sea level. The sky was overcast and the visibility was good. The descent profile was shallow because the proposed landing area was one valley over and only about 2000feet below. While pilots may use the rpm beep trim switch to prevent engine over-speed during a steep descent, the accident flight did not require its use. In addition, when the pilot noted the loss of rotor rpm, he checked, by pushing forward on the beep trim switch, that the rpm was beeped high. The helicopter's gross weight was about 8200pounds, 3000pounds below the maximum gross weight. The lack of engine response occurred in a flight regime where the pilot's options were limited. There was insufficient altitude and air speed to put manual governor operations into effect. After the accident, fuel was drained from the fuel cell sumps; the fuel was clean and appropriate. The helicopter was later inspected in a controlled environment. All of the engine controls were intact and functioned properly, and the rpm trim was beeped all the way up. The airframe fuel filters were inspected and contained clean fuel. The torque control unit (TCU) was removed and inspected, and, based on a simple test, the TCU ports appeared to be closed. The oil magnetic chip plugs from the engines and combining gearbox (C-box) were inspected. Only the C-box centre plug contained metal chips. The C-box oil filter was removed and inspected; it contained silver and iron particles. The engine power turbine (PT) sections and the C-box were rotated backwards, and they appeared to turn normally. When the drive shaft was rotated forward, the PT sections did not turn, indicating that the free wheel units (one-way clutches) were functioning. The Twin-Pac - the two engines and the combining gearbox - was removed from the helicopter. The TCU was reinstalled, and the Twin-Pac was installed in an approved engine test facility, and several test runs were carried-out. The engines did not operate as they should. Both engines' rpm oscillated and the engines would not accelerate on demand. When the engines were operated in a manual mode, as opposed to automatic governor mode, they operated normally. The TCU was disconnected and the problem remained. The power turbine (Nf) governors were replaced with similar governors that had about 400hours of in-service use. With these installed, the engines operated normally, with or without the TCU connected. It was determined that the Nf governors on the accident helicopter were not rigged normally, in that their control arms were statically positioned at 74 and 73, as opposed to standard positions of about 85 or 90. The accident Nf governors were reinstalled on the Twin-Pac with their control arms set to 90. The engines were run again, and they operated normally. The substitute Nf governors were then installed again, but rigged at 74 and 73. When run in this configuration, the rpm oscillated and the engines would not accelerate on demand. This time, however, oscillations occurred at a higher rpm. In this configuration, the engines ran normally when run independently. The accident governors were bench tested and they did not function properly. With the control arms set at 75, both governors bled Pgair at 11.9inches Hg when Pg air should bleed at between 6.8 and 7.8inches of Hg. The governors were disassembled and inspected. Both had in-service wear, both showed wear on the bushings on which the control arms pivot, and there were indications of uneven seating of the port valves. The wear was excessive, given that the Nf governors had only 823hours of time in service and the overhaul life is 4500hours. Returning these governors to service would have required that some parts be replaced. It is noteworthy that many similar governors used in this application are removed from service at about 1600hours because of fluctuating torque, and that they have shown similar wear. A review of the helicopter's maintenance records and discussion with the operator's personnel revealed that maintenance personnel had replaced the TCU in an effort to resolve torque fluctuations. A replacement TCU had been requested by the operator, but none was readily available. Pratt Whitney Canada suggested using a new version of the TCU that was approved and being used on -3B engines, with good results, by another operator. The company provided instruction on how to install the new TCU. However, the new TCU had not gone through the approval process for use on the -3DF engines installed on the accident helicopter. This new TCU was different from the one removed in that it had its Pg constant bleed ports capped and, therefore, blocked. This affected the Nf (turbine speed) on both engines, so the control arms were adjusted to bring the Nfs into the normal range. While this rigging change was required, the change was not incorporated in any of the approved maintenance manuals. The new TCU stabilized the torque somewhat, and it remained installed for about 150hours of flight time before the accident. The Twin-Pac had been installed for about 823flight hours. The engines and the governors were new when they were installed. A well-known, approved overhaul and repair facility reported that the average time on these types of governors when they are received for repair due to wear was about 1600hours. Pratt Whitney Canada and Bell Helicopters Textron worked with the industry in autumn2002, in an effort to resolve problems with TCUs. It was found that the Nf governors performed better and torques were more stable without the TCUs. Bell Helicopters Textron did not approve the removal of the TCUs, but worked with Pratt Whitney Canada to approve a TCU with the constant bleeds blocked. These TCUs, with blanked off (Pg) orifices, have been in limited use since January2003 (Pratt Whitney Service Bulletin5463) for the PT6-3B engines with specific, serial-numbered, reduction gearboxes. Pratt Whitney Canada worked on approvals for these modified TCUs for all Twin-Pac installations in spring2005. Except for the TCU, the engine controls on the Bell212 operate independently. The engines deliver power to the main-rotor transmission through the combining gearbox. In the event one engine loses power, the other engine can deliver its full power for an emergency landing. The TCU is a simple apparatus that limits combined engine torque and, in some models, balances torque. The TCU's purpose is to protect the helicopter from over-torque when both engines are operating. Anecdotal information suggests that many operators adjust their TCUs to allow pilots a margin of error, preferring to change drive components in the event of an over-torque than have their helicopters crash with unabused drive components. During the time of the Twin-Pac tests, there was no accumulation of metal on the magnetic chip plugs. Anecdotal information suggests that metal (silver and iron) particles can be expected in the combining gearbox oil during normal operation.